How can the max characters for the message field in the csv for a data source be 65,536 when the max characters for a field in a csv file is only 32,767? Also, the use case I’m working on is for faxes where it’s not uncommon to have > 30 pages (so far greater than 65,536) - does that mean comms mining is not a good solution for extracting data from these completely unstructured PDFs, or is the another way comms mining can still be used?